68 RHODOSPERMEjE. 



to our observation. One sort of fruit, the tetraspore, 

 though extremely various in position, is uniform in structure 

 throughout the whole sub-class. It consists invariably of a 

 membranous, sub-gelatinous sac or perispore, containing a 

 mass of red colouring matter that separates, at maturity, into 

 four parts or sporules ; sometimes by a transverse division 

 (zoned or annular] ', sometimes by two cross lines into four 

 equal parts (cruciate] ; and sometimes by triradiate lines into 

 four unequal parts (ternately-parted]. This little body is 

 very rarely found in a proper conceptacle or capsule, as we 

 should expect the representative of a seed to be. In the co- 

 rallines, in the Australian genus Ctenodus, and in a few other 

 instances, we do find tetraspores enclosed in hollow cases. 

 In many others they are found in pod-shaped bodies called 

 stichidia, as in Rytiphl&a, Dasya, Plocamium, &c. ; these 

 stichidia being sometimes formed by alteration of portions of 

 branchlets or leaves, and sometimes independently developed 

 from definite points of the frond. In others the tetraspores 

 are naked (as in Callithamnion), scattered over the sides, or 

 fixed at the tips of the branchlets. But in by far the greater 

 number of cases these little bodies are immersed in the sub- 

 stance of the branches or leaves, making no external show, 

 except that the parts where they congregate are of darker 

 colour than the rest ; and they must be sought by careful ex- 

 amination and dissection under a lens. In these latter cases 

 they appear to be formed either from the cells of the external 

 coat, or from those immediately beneath the surface-cells. 

 If we give due weight to the evidence derived from position, 

 it appears to me that the weight of that evidence would favour 

 the supposition that tetraspores were gemmules, and not true 

 spores. And such is the opinion advocated by Mr. Thwaites, 

 contrary to that of Decaisne, of J. Agardh, and of almost all 

 modern writers of repute. In cases where they occur dis- 

 persed through the frond, one can hardly conceive their 

 formation by a sexual process ; and in such plants as Cal- 

 lithamnion they clearly originate in the alteration of one 

 of the cells of the ramuli. In this genus also we fre- 

 quently find them viviparous, or converted into innumerable 

 graniform cells, strung together. Such bodies have been 

 called antheridia, but, I think, without sufficient warrant. 

 To me they have always appeared to be metamorphic tetra- 

 spores. Again, the advocates of the opposite view, who 

 regard the tetraspore as a true spore, appeal to its perfectly 

 regular structure, uniform through the whole sub-class, and 



